Social media has made it easier than ever to publish pictures on the internet. Investigators now have to determine how to deal with pictures, posted to these social media websites, which become evidence in crimes. Traditional methods of authenticating the origin of pictures are not as effective because the processing applied by the social media websites eliminates or obscures much of this information. This thesis proposes several methods of image attribution that utilize popular image authentication techniques and presents test data to illustrate their use. Chapter 1 introduces to the topic of Image Attribution. Chapter 2 provides brief overview of the background analysis techniques that will be used. Chapter 3 describes how the test data that was taken and why. Chapter 4 proposes three different methods for answering three different forms of image attribution. Chapter 5 tests the reliability of the proposed methods with the cell phone image database uploaded to three separate social media websites. Chapter 6 provides a conclusion based on the results test results. Chapter 7 proposed future research in the field.
College of Arts and Media
Copyright of the original work is retained by the author